RapidChart MCP Server

# RapidChart MCP Usage Examples ## Basic Examples ### 1. Generate Class Diagram ``` User: "Generate a class diagram from this code: class User { id: number; name: string; email: string; password: string; } class Post { id: number; title: string; content: string; author: User; createdAt: Date; }" Response: ✅ Diagram created successfully! Title: User and Post Classes Type: class URL: https://rapidchart.com/editor/abc-123 ``` ### 2. Generate from File ``` User: "@src/models/user.py Generate a class diagram from this file" → Cursor reads the file automatically → Calls create_diagram with file contents → Returns diagram URL ``` ### 3. Generate ER Diagram ``` User: "Generate an ER diagram from this SQL schema: CREATE TABLE users ( id SERIAL PRIMARY KEY, email VARCHAR(255) UNIQUE, name VARCHAR(100) ); CREATE TABLE posts ( id SERIAL PRIMARY KEY, user_id INTEGER REFERENCES users(id), title VARCHAR(255), content TEXT );" Response: ✅ Diagram created successfully! Type: er URL: https://rapidchart.com/editor/def-456 ``` --- ## Advanced Examples ### 4. Use Specific AI Model ``` User: "List my available models" Response: 📊 Available AI Models: ✅ [10] GPT-5 mini (unlimited credits) ✅ [11] GPT-5.2 (using your API key) ❌ [12] GPT-5 nano (0 credits) User: "Generate class diagram with model 11 from src/models.py" → Uses GPT-5.2 specifically ``` ### 5. Update Existing Diagram ``` User: "List my diagrams" Response: [abc-123] User Models (class) [def-456] Database Schema (er) User: "Update diagram abc-123 with new code from src/models.py" → Fetches old diagram → Passes as context to AI → Regenerates with awareness of old structure ``` ### 6. Organize in Workspaces ``` User: "List my workspaces" Response: 🏠 Personal [1] My Workspace 👥 Team [5] Company Project User: "Generate diagram in workspace 5 from src/api.py" → Creates diagram in team workspace ``` ### 7. Use Folders ``` User: "List folders in workspace 5" Response: [10] Backend [11] Frontend [12] Database User: "Generate ER diagram in workspace 5, folder 12 from schema.sql" → Creates diagram in: Company Project / Database / ``` --- ## Agent Mode Examples ### 8. Multi-File Analysis ``` User: "Analyze all files in src/models/ and generate a comprehensive class diagram" → Cursor agent: 1. Lists files in src/models/ 2. Reads each file 3. Combines code 4. Calls create_diagram with all code 5. Returns URL ``` ### 9. Code Change Detection ``` User: "I updated src/models/user.py, update the User Models diagram" → Cursor agent: 1. Reads updated file 2. Searches for "User Models" diagram 3. Calls update_diagram with new code 4. Returns updated URL ``` ### 10. Architecture Diagrams ``` User: "Generate AWS architecture diagram for this infrastructure: - API Gateway - Lambda functions - DynamoDB - S3 bucket - CloudFront CDN" → Creates AWS architecture diagram with proper icons ``` --- ## Workflow Examples ### 11. Development Workflow ``` # Day 1: Initial design User: "Generate class diagram from src/models/" → Creates diagram abc-123 # Day 15: Code evolved User: "Update diagram abc-123 with current code from src/models/" → Updates with new classes/relationships # Day 30: Refactoring User: "Update diagram abc-123, I refactored the User class" → Regenerates with context ``` ### 12. Documentation Workflow ``` # Generate diagrams for docs User: "Generate sequence diagram for authentication flow" User: "Generate ER diagram for database schema" User: "Generate C4 diagram for system architecture" # Export URLs for documentation User: "List all my diagrams from today" → Copy URLs into docs ``` --- ## Tips & Best Practices ### ✅ DO: - Use descriptive titles for diagrams - Specify diagram type explicitly - Check available models first with `list_models` - Use workspace/folder organization for large projects - Update diagrams when code changes significantly ### ❌ DON'T: - Don't paste huge files (>1000 lines) - summarize or split - Don't regenerate diagrams for tiny changes - Don't forget to specify diagram type (defaults to "general") - Don't use expired tokens --- ## Diagram Types Reference | Type | Description | Best For | |------|-------------|----------| | `general` | Generic flowchart | Processes, workflows | | `class` | UML class diagram | OOP code, data models | | `er` | Entity-relationship | Database schemas | | `sequence` | Sequence diagram | API flows, interactions | | `usecase` | Use case diagram | System requirements | | `userflow` | User flow | UI/UX flows | | `c4` | C4 architecture | System architecture | | `aws` | AWS architecture | Cloud infrastructure | | `azure` | Azure architecture | Azure cloud | | `google` | GCP architecture | Google Cloud | --- ## Common Patterns ### Pattern 1: Code-to-Diagram Pipeline ```bash # In your CI/CD rapidchart-mcp generate \ --code "$(cat src/models/*.py)" \ --type class \ --title "System Models" \ --output diagram-url.txt ``` ### Pattern 2: Documentation Generator ```python # Script to generate all diagrams diagrams = [ ("src/models/", "class", "Data Models"), ("db/schema.sql", "er", "Database Schema"), ("docs/flows.md", "sequence", "API Flows") ] for path, type, title in diagrams: code = read_file(path) url = create_diagram(code, type, title) print(f"{title}: {url}") ``` --- ## Troubleshooting Common Issues ### Issue: "Diagram looks wrong" **Solution:** - Enable guidelines: `RAPIDCHART_GUIDELINES=true` - Enable thinking: `RAPIDCHART_FEW_PROMPTS=true` - Be more specific in your prompt - Try a different AI model ### Issue: "Too slow" **Solution:** - Disable few_prompts for faster generation - Use GPT-5 mini instead of GPT-5.2 - Reduce code size (summarize or split) ### Issue: "Wrong diagram type" **Solution:** - Explicitly specify type: "Generate CLASS diagram..." - Check default type in config - Agent might misinterpret - be explicit --- For more examples, see the [README](../README.md) and [Cursor Setup Guide](../CURSOR_SETUP.md).